3.11) Robotics, artificial intelligence and expert systems
Robotics
Input devices: for example, camera, sensors, microphones
Input device: Any hardware device that sends data to the computer, without any input devices, a computer would only be a display device and not allow users to interact with it
Output devices: for example, claws, wheels, motors, relays, speakers
any device that outputs information from a computer, it can output it in either a auditory or visual format.
Robot, android, cyborg
Robot: The most common type of robot is the industrial robot. They are used in factories, it looks like a human arm, and it also has joints.
The jobs that they do:
lift heavy items into from place to place
assemble parts together to create things
join parts together using glue, or by welding (melting metal)
paint things
Android: A robot that is designed to look like and act like a human
Cyborg:A human who has certain physiological processes aided or controlled by mechanical or electronic devices.
Sensors: for example, heat, proximity, magnetism, light, humidity, pH
Artificial intelligence
YouTube Clip (Stanford University lecturer on AI): Control, Perception, Human Brain processes
Artificial intelligence versus computational intelligence
Artificial intelligence is the intelligence of software or hardware and us defined often by that the system is able to perceive its environment and be able successfully complete tasks in that environment without being controlled by a human.
Computational intelligence is similar to Artificial Intelligence, but computational intelligence works in a different way. A system using Computational Intelligence doesn’t need sensors and it takes a mathematical approach to solving complex tasks. Computational Intelligence works from numbers to interpret symbols around it and meanwhile AI uses symbols to interpret the environment around it.
Man or machine: Turing test, CAPTCHA (completely automated public Turing test to tell computers and humans apart)
Turing Test is a test to determine whether a computer is able to simulate the behavior of humans to be passed off as a human.
CAPTCHA is a test designed to tell humans and computers apart. Usually the test involves a piece of text which has been distorted or modified so that a computer will struggle to read what it says and humans should be able to see what it is saying. CAPTCHA prevents email spam and unnecessary load on systems which have a large amount of traffic.
Capabilities and limitations: for example, learning to identify human emotions, evaluation of living things and machines (intuition, prior knowledge, judgment)
AI techniques: searching, pattern recognition, heuristics, machine learning
Fuzzy logic, set theory
Machine learning: can machines become independent?
Machine learning is the method of teaching computers to improve behaviors functions, depending on the data given. This data can be reading from sensors placed on the computer. Machine learning is essentially applied pattern recognition.
Natural language communication and translators
Natural language communication is the capability for a machine to interact with a person using everyday language. Natural language processing requires the machine to understand the natural language input by the human. Natural language translators basically translate language to another automatically. Natural language translation is extremely painstaking to solve properly as the machine needs to properly understand grammar, semantics and real-world facts, possibly even accents if oral.
Neural networks: similarity to biological systems
Artificial neural networks are a group of artificial neurons that use a biologically inspired approach to giving computers intelligence. Neural networks modify parameters in response to expectations. In theory, neural networks can perform any behaviour as they can approximate any function to excellent precision if time is negligible.
Pattern recognition is the machines capability to understand real attributes using sensors. In AI, pattern recognition is used to input data into the computer.
Processing and storage requirements
Expert systemsis a computer system that can match decision-making abilities of a human expert
Collection, creation and maintenance of knowledge base Knowledge base is a type of database for knowledge management. It's an information repository that provides a means for information to be collected, organized shared, searched and utilized. It can be machine-readable or be used for human use.
Creation of inference engine, inference rule (“if–then” rules), chaining, suitable domains for expert systems Inference engine is a computer program that tries to find answers from a knowledge base. Essentially this is a brain of an expert system and its purpose is to find new conclusions by using logical reasoning. There are special types of reasoning engines which can use more general types of reasoning. Inference rulecan be a rule or function which takes an idea and then returns to a conclusion based on this idea. Chaining:
Forward Chaining is one of the two main methods of reasoning when using inference rules. It starts with avaible data and uses inference rules to extract more data from the end user until a goal is reached. An inference engine uses forward chaining to search the inference rules until it finds one where the antecedent (IF rule) is true.
Purpose of an algorithm within expert systems: for example, fault finding, product development
Notes: What is a Robotics & Artificial intelligence?
1. - where they are today? where we will be going in the future? - idea: build a worker - robots in Australia: mining, mapping, fly robot aircraft's-- uses sensors - what it is: a computer that connects to the real world- through: action, learning, perception
The theory behind robotics, robots don't process data and information in the same way as we do. Humans think and see objects in a different way, since we use our senses. When humans think of a lemon we think of its taste and how it looks, but robots can't see it in the same way, so they have to see thousands of examples of lemons to learn that it is a lemon. Robots do things in steps, they analyse the things around them and after they are confident that they should do a thing. Robots rely on their sensors and the algorithms that tell them what to do. Algorithms are critical in robots and they are key to making robots function. Robots are being taught how to learn and they are even developing a fully functioning memory.
What is AI -Deep blue by IBM beat the world champion at Chess -Machines can learn and robots are learning in the same way as small children do. -Machine Learning Algorithms -Make a system learn and choose its own goals. -Humanoid robots are physically able to grasp and control objects -Robots have a massive stream of data and the act of picking up a glass is very complex for robots to learn. -MobeE, breaks down complex problems and makes them into more manageable tasks. -MobeE even knows if it has succeeded or failed in picking up the object, and it can sense whether the object is standing or has toppled. - Robots use detectors and locators to be able to pick up the object. - Robots also use filters filter out the parts of the image they don't need to think about. - It can even differentiate between whether an action was deliberate or reflexive. - NES controls the actions that should be executed and it will learn from mistakes and therefore make the robot perform the action more efficiently is.
3. create software and different algorithms for speech recognition and visual recognition. the idea: A human brain does not need a lot of algorithms for perception --> one single learning program for the robot... we have not yet found that program
neural network: like our brains neurons are connected, a robot has a neural network
3.11) Robotics, artificial intelligence and expert systems
Robotics
- Input devices: for example, camera, sensors, microphones
- Input device: Any hardware device that sends data to the computer, without any input devices, a computer would only be a display device and not allow users to interact with it
- Output devices: for example, claws, wheels, motors, relays, speakers
- any device that outputs information from a computer, it can output it in either a auditory or visual format.
- Robot, android, cyborg
- Robot: The most common type of robot is the industrial robot. They are used in factories, it looks like a human arm, and it also has joints.
- lift heavy items into from place to place
- assemble parts together to create things
- join parts together using glue, or by welding (melting metal)
- paint things
- Android: A robot that is designed to look like and act like a human
- Cyborg:A human who has certain physiological processes aided or controlled by mechanical or electronic devices.
- Sensors: for example, heat, proximity, magnetism, light, humidity, pH
YouTube Clip (Stanford University lecturer on AI): Control, Perception, Human Brain processesThe jobs that they do:
Artificial intelligence
Artificial intelligence is the intelligence of software or hardware and us defined often by that the system is able to perceive its environment and be able successfully complete tasks in that environment without being controlled by a human.
Computational intelligence is similar to Artificial Intelligence, but computational intelligence works in a different way. A system using Computational Intelligence doesn’t need sensors and it takes a mathematical approach to solving complex tasks. Computational Intelligence works from numbers to interpret symbols around it and meanwhile AI uses symbols to interpret the environment around it.
Turing Test is a test to determine whether a computer is able to simulate the behavior of humans to be passed off as a human.
CAPTCHA is a test designed to tell humans and computers apart. Usually the test involves a piece of text which has been distorted or modified so that a computer will struggle to read what it says and humans should be able to see what it is saying. CAPTCHA prevents email spam and unnecessary load on systems which have a large amount of traffic.
- Capabilities and limitations: for example, learning to identify human emotions, evaluation of living things and machines (intuition, prior knowledge, judgment)
- AI techniques: searching, pattern recognition, heuristics, machine learning
- Fuzzy logic, set theory
- Machine learning: can machines become independent?
Machine learning is the method of teaching computers to improve behaviors functions, depending on the data given. This data can be reading from sensors placed on the computer. Machine learning is essentially applied pattern recognition.- Natural language communication and translators
Natural language communication is the capability for a machine to interact with a person using everyday language. Natural language processing requires the machine to understand the natural language input by the human. Natural language translators basically translate language to another automatically. Natural language translation is extremely painstaking to solve properly as the machine needs to properly understand grammar, semantics and real-world facts, possibly even accents if oral.- Neural networks: similarity to biological systems
Artificial neural networks are a group of artificial neurons that use a biologically inspired approach to giving computers intelligence. Neural networks modify parameters in response to expectations. In theory, neural networks can perform any behaviour as they can approximate any function to excellent precision if time is negligible.- Pattern recognition: OCR (optical character recognition), image analysis, speech recognition, speech synthesizers
Pattern recognition is the machines capability to understand real attributes using sensors. In AI, pattern recognition is used to input data into the computer.
- Processing and storage requirements
- Collection, creation and maintenance of knowledge base
- Creation of inference engine, inference rule (“if–then” rules), chaining, suitable domains for expert systems
- Expert systems, knowledge base, knowledge engineer, expert system shells, inference engine, domain, common-sense knowledge
- Purpose of an algorithm within expert systems: for example, fault finding, product development
Notes:Expert systems is a computer system that can match decision-making abilities of a human expert
Knowledge base is a type of database for knowledge management. It's an information repository that provides a means for information to be collected, organized shared, searched and utilized. It can be machine-readable or be used for human use.
Inference engine is a computer program that tries to find answers from a knowledge base. Essentially this is a brain of an expert system and its purpose is to find new conclusions by using logical reasoning. There are special types of reasoning engines which can use more general types of reasoning.
Inference rule can be a rule or function which takes an idea and then returns to a conclusion based on this idea.
Chaining:
Forward Chaining is one of the two main methods of reasoning when using inference rules. It starts with avaible data and uses inference rules to extract more data from the end user until a goal is reached. An inference engine uses forward chaining to search the inference rules until it finds one where the antecedent (IF rule) is true.
What is a Robotics & Artificial intelligence?
1.
- where they are today? where we will be going in the future?
- idea: build a worker
- robots in Australia: mining, mapping, fly robot aircraft's-- uses sensors
- what it is: a computer that connects to the real world- through: action, learning, perception
The theory behind robotics, robots don't process data and information in the same way as we do.
Humans think and see objects in a different way, since we use our senses. When humans think of a
lemon we think of its taste and how it looks, but robots can't see it in the same way, so they have to see thousands of examples of lemons to learn that it is a lemon. Robots do things in steps, they analyse the things around them
and after they are confident that they should do a thing. Robots rely on their sensors and the algorithms that
tell them what to do. Algorithms are critical in robots and they are key to making robots function. Robots are being
taught how to learn and they are even developing a fully functioning memory.
What is AI
-Deep blue by IBM beat the world champion at Chess
-Machines can learn and robots are learning in the same way as small children do.
-Machine Learning Algorithms
-Make a system learn and choose its own goals.
-Humanoid robots are physically able to grasp and control objects
-Robots have a massive stream of data and the act of picking up a glass is very
complex for robots to learn.
-MobeE, breaks down complex problems and makes them into more manageable tasks.
-MobeE even knows if it has succeeded or failed in picking up the object, and it can sense whether the
object is standing or has toppled.
- Robots use detectors and locators to be able to pick up the object.
- Robots also use filters filter out the parts of the image they don't need to think about.
- It can even differentiate between whether an action was deliberate or reflexive.
- NES controls the actions that should be executed and it will learn from mistakes and therefore make the robot
perform the action more efficiently is.
3.
create software and different algorithms for speech recognition and visual recognition.
the idea: A human brain does not need a lot of algorithms for perception
--> one single learning program for the robot... we have not yet found that program
neural network: like our brains neurons are connected, a robot has a neural network
Useful links:
HL extension: